Skip to content
Back to skills

Review Contract

ASecurity

Review a commercial contract against a configurable playbook — flag deviations, generate redlines, and provide business-impact analysis. Covers M&A, services / SOW, license / IP, employment, distribution, supply, JV, settlement, and other commercial agreements. Use for clause-by-clause analysis against standard positions, prioritized redlines, and fallback positions.

  • 10 stars
  • 0 votes
  • 0 copies
  • 2 views
  • Added August 31, 2026
ai-agentsgosecurity

Works with

  • cli

Security analysis

A100/100

Scanned August 31, 2026

npx -y skills add WenzhuoXu/lawgent --skill review-contract --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Review Contract?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Review Contract
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/wenzhuoxu-review-contract/badge)](https://www.skillsdirectory.com/skills/wenzhuoxu-review-contract)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: review-contract
description: Review a commercial contract against a configurable playbook — flag deviations, generate redlines, and provide business-impact analysis. Covers M&A, services / SOW, license / IP, employment, distribution, supply, JV, settlement, and other commercial agreements. Use for clause-by-clause analysis against standard positions, prioritized redlines, and fallback positions.
argument-hint: "<contract file or text>"
allowed_connectors: [ecfr_search, courtlistener_search, eurlex_search, "pkulaw_*", flk_npc_search]
rag_collections: [general]
---

# /review-contract — Contract Review

Produce the analytical bundle the orchestrator turns into a redline +
business-impact memo. The skill identifies deviations from the playbook,
not user-facing prose.

**Not legal advice.** Drafts must be reviewed by qualified counsel before
counter-signing.

## Output Contract (binding)

Same as `/playbook/general_playbook.md` §4 and §5:

````
## Findings
- [self-contained claim, ONE sentence, ≤60 English words, inline pinpoint `[Source, art./§/p.]`] (label: treaty | regulation | case | guidance | local-law | best-practice | risk | drafting)
- ...

## Out of scope
- ...

## Sources
- [source name] — [URL or local path] — supports findings #X[, online-checked: yes/no, pinpoint: confirmed/unavailable]
````

Hard ceiling **≤ 2000 characters**. No redline tables or "must-have /
should-have / nice-to-have" tier blocks in the bundle — those are the
orchestrator's job at authoring time.

## Inputs

- **Contract** (file attachment preferred; falls back to pasted text).
- **Which side** the client is on (party A / party B / mutual).
- **Contract type** (sale, services, licence, employment, distribution,
  M&A, NDA, supply, JV, settlement, lease).
- **Jurisdiction** + governing-law clause.
- **Stage** (term-sheet check, first-mark, redline response, closing).
- **Risk appetite** override if it differs from the playbook default.

## Workflow

1. **Read the playbook**: `/playbook/general_playbook.md` for global
   defaults; the active domain pack's playbook for domain positions; any
   user-supplied client-specific playbook.
2. **Read the contract structure** — parties, scope, term, fees,
   warranties, indemnities, limitation of liability, IP, confidentiality,
   data-protection, change of control, termination, dispute resolution,
   governing law.
3. **Clause-by-clause analysis**: surface one Finding per clause that
   deviates from the playbook, with the deviation type (missing /
   defective / one-sided / unenforceable / non-market) and a one-sentence
   fallback position.
4. **Cross-jurisdictional traps** — enforceability in PRC vs US vs EU
   (e.g. liquidated damages, exclusion of consequential loss, choice of
   forum, governing-language).
5. **Risk hooks** — flag any deviation that triggers an escalation per
   `/playbook/general_playbook.md` §6 (regulator-disclosure carveout,
   sanctions, data localisation, state secrets, privilege).
6. **Sanity check** per general playbook §3.

## Clause buckets (generic)

Parties + recitals; definitions; scope / SOW; fees + invoicing; terms +
termination; warranties; indemnities; limitation of liability;
confidentiality + NDA; data protection + DPA; IP ownership + licences;
service levels; security + audit; insurance; change of control;
assignment + novation; force majeure; dispute resolution + governing law;
boilerplate (notices, severability, entire-agreement, counterparts).

## Pointers

- Authority hierarchy + citation: `/playbook/general_playbook.md` §0 + §2.
- Sanity-check + label set: general playbook §3 + §4.
- Universal risk defaults (data, sanctions, privilege, state secrets,
  regulator-disclosure): general playbook §6.
- When the **aviation** pack is active, also apply
  `/domains/aviation/overlays/review-contract.md` for aircraft / charter /
  maintenance / OEM agreements.

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.